SPY Broke 740, SPCX Flushed To $110, Brent Ripped Past $100 — Is The Memory And SpaceX Trade Over, And Why Is Money Still Bid With Oil?

The 740 line went. That was the master pivot I've been running for a week and the tape took it out on the Google/Tesla after-hours hangover into a Brent tape that ripped past $100. S&P is testing the 735 shelf as I write this — that's the next line that matters, and where the question 'flush or not' actually gets answered.

SPCX printed around $110-116 today. That's down roughly 49% from the June 16 peak at $225.64 and 13.5% below the $135 IPO price. A trillion-plus of paper market cap is off the tape in six weeks. The question the desk is asking me: is the SpaceX trade over, or is $110 the bottom of the barrel?

And the weird part everyone is trying to reconcile: money is still bid — cash rates high, credit tight, gold firm, dollar firm — and oil is bid at the same time. Normally you pick one. Today's tape says you pay for both. Here's why, whether SNDK actually flushed, and the levels I'm running into tomorrow.

Short version: 740 broke, 735 has not confirmed. SPY sliced the 740 pivot on the cash open after Alphabet's capex-hike sold off in after-hours and Tesla's EPS miss ran the futures overnight. The tape is testing the 735 shelf in the afternoon session as Brent broke $100. That is a break-and-test, not a flush. A daily close below 735 is the flush signal — and that opens 728 fast, 720 on any follow-through.

The line I'm using: 735 confirms with a daily close below it plus VIX above 22 plus the 10-year holding 4.65%+. If you get all three, you are not looking at a shakeout, you are looking at a regime shift where the AI capex bid is no longer big enough to offset the macro tape and you play for 720-into-712 next week.

The line that says 'not flushed yet': a reclaim of 740 into the close, VIX back under 20, and any sign the 10-year is fading back under 4.60% on a bid-to-cover story out of Treasury. In that case today is a scary-looking single-day shake and 745 gets tagged again by Monday. That is not my base case right now — but it is the only setup that keeps the bulls honest.

The honest read: the tape is one weak Microsoft print or one Hormuz headline away from a real 735 close. It is also one dovish Fed leak or one crude reversal away from reclaiming 742. That is where we are. Don't sell the flush on the break — sell it on the confirmation.

Every desk is asking the same thing because they see the SPY chart and assume memory got hit. SNDK closed the prior session at $1,599.27, printed an intraday high of $1,628.40, and is holding well above the pre-print base. From the $1,340 after-hours entry, that is a clean win still on the tape. Memory did not flush. Memory sold two ticks and bid.

Why: the Alphabet capex raise is a delayed order for HBM and enterprise NAND. The tape figured that out overnight. Every hyperscaler capex dollar rolls downstream into memory 6-18 months later, and the sell-side is walking numbers up, not down. That is why SNDK is not tracking SPY on a session where the index cracked 740.

The setup I'm actually stalking: a real range compression — a 5-8 session tape where SNDK ranges between $1,500 and $1,650, fast money rotates out, and the character resets. That is the reload window. A flush-flush — the kind that gets talked about — would look like $1,350 with a headline that HBM 4 got pushed. Nothing on the wire says that. So no, SNDK has not flushed. I am flat from the last exit and watching the range, not chasing the tape.

The tell for a real memory flush, when it comes: a Microsoft or Amazon print that soft-guides on cloud growth against a raised capex number. That is the 'peak margin' bear thesis lighting up. Until that shows up on the tape, treat every red day in SNDK as noise around a bid, not the start of a trend break.

Facts first. SPCX printed $115.26 on July 23, down 6.7% on the session, down about 49% from the June 16 all-time high of $225.64, and down 13.5% from the $135 IPO price. That is over $1T of paper cap erased in six weeks. Six-day losing streak. New all-time low intraday. This is the ugliest tape any newly-public mega-cap has printed in a decade.

Three things are hitting simultaneously and the market is pricing all of them at once: (1) the July 16 Starship 13 abort — 4 of 33 booster engines cut one second before liftoff — which put an actual physical risk premium back into the Starlink revenue trajectory; (2) Morgan Stanley flagging that xAI capex demand on SpaceX capacity could balloon to $120B, which the market read as 'more dilution, more capex-ahead-of-revenue'; (3) the standard post-IPO lockup expiration overhang combined with insiders and pre-IPO investors trimming into any strength. That is a three-vector sell-off and until one of them clears, the bid is thin.

Is it over? No. Not at $110. Here is why. The float is still light relative to the natural demand for SpaceX exposure globally, and Starlink alone is a real cash-flowing business — revenue run rate is comfortably above $12B annualized with ARR growth still in the high 30s. That business does not deserve a sub-$135 tape once the Starship overhang resolves. The setup you actually want: one clean successful Starship integrated flight test, one guidance update that shows Starlink margin holding despite xAI capacity carve-outs, and the trapped longs from the IPO capitulate into a single-session low. That is the bottoming pattern.

How I'd trade it: I am not touching SPCX for a long here on the break of $110. I want to see a session that trades $105 handle, closes green, and holds the low the next morning. That is the reversal candle. Until then it is a knife with a story attached. Not everything that is down 49% is a buy; some things are down 49% and going to $80 before they see $150 again. Respect the tape.

This is the question the desk keeps chewing on. Money — the front-end curve, cash yields, the dollar, credit spreads — is bid. Oil is bid. Historically that combo is unusual because higher oil normally comes with reflation, and reflation normally means selling front-end paper and selling the dollar. Not this time. Here's the read.

One — the oil bid is a geopolitical option, not a growth signal. Brent through $100 today is priced on the Houthi tanker attacks in the Red Sea, Trump threatening 'a massive attack' on Iranian infrastructure, and the twelfth night of US-Iran exchange. This is a supply-risk premium, not demand. The market is buying the option that a real barrel stops moving. In that world, you also want short-duration paper and cash, because a supply shock is stagflationary — it's tax on growth, not fuel for it. So oil goes up and cash yields stay bid together. That is coherent, not a contradiction.

Two — the equity bid inside a 'money-bid' regime is narrow and it is being paid for by AI cash flows. The Alphabet print showed Cloud growing 82% with a $92B capex plan. That is real earnings power that discounts higher rates. So the S&P can crack 740 on a Google/Tesla hangover and still have a natural bid at 735 because the earnings compounder underneath is intact. What gets sold is junk beta, long-duration hope, unprofitable growth. What gets bought is Big Tech cash flow, energy, defense, gold. That is exactly what you see in the index composition today.

Three — the dollar and gold both bid at the same time is the tell that the world is buying insurance while staying long the equity book. That is a fragile equilibrium. It works until it doesn't. The break point is a supply outage in Hormuz or a soft Microsoft cloud print. Either one, and the fragile equilibrium tips and the money-and-oil pair breaks in favor of oil-only, with equities catching down.

The one-line frame: money is bid because a supply shock is stagflationary, oil is bid because Hormuz is a real option, and equities are bid narrowly because AI cash flows are earning the discount rate. Trade the frame. Don't fight it and don't try to explain it away.

SPY: 740 broke. 735 is the pivot. Daily close below 735 with VIX > 22 opens 728 fast, then 720. A reclaim of 740 into the close with VIX back under 20 puts 745 in play by Monday. I'm hedged with QQQ puts, not flat outright. The hedge comes off on a 740 reclaim.

SNDK: $1,500 is the base. $1,650 is the range top. Below $1,480 with volume is the character break — that is the memory flush the tape has been waiting for and I'd be all over it long from $1,400-1,420 if it prints. Above $1,650 with follow-through is a fresh leg toward $1,800. Between the lines: patience, no chase.

SPCX: $110 is the shelf. $105 is the panic level. Setup for a long entry is a $105 tag that closes back above $110 and holds the low the next day. Not touching it until that pattern shows. Above $130 on volume with a Starship win is the trend reversal that reopens $150.

Brent (BZ=F): $100 is the line. Above $100 daily close and the front end reprices harder and every risk asset re-rates. Below $96 on a de-escalation headline and the option premium drains fast — that is the crude fade I've been stalking for a week.

10-year: 4.65% is the shelf. Above 4.70% and equity multiples take real damage. Under 4.55% and the front-end unwind is on and 745 SPY prints in a session.

VIX: 22 is the trigger. Under 20 and this is noise. Above 22 sustained and the S&P 735 close probability doubles.

Long COIN — the stablecoin flywheel does not care about Brent or Hormuz, and the $200-plus target is still the number I'm playing for.

Flat SNDK, watching for the range-compression reload. Memory did not flush and I am not chasing the tape up.

No SPCX exposure. I want the $105 tag and the reversal candle before I touch it. Down 49% is not the same as bottomed.

Hedged with QQQ puts around the 740 break. That is the risk-off expression I want, not outright flat. The hedge comes off on a 740 reclaim into the close.

No new crude exposure. Long XLE only on an actual Hormuz outage, XLE fade only on a real de-escalation headline. Flat and watching. This is not the tape to chase.

Gold — small starter is fine as insurance if you don't already own it. Dollar strong plus gold strong is the tell I'm respecting.

740 broke. 735 has not confirmed. SNDK did not flush. SPCX is ugly but not bottomed. Brent through $100 is a Fed problem, not a supply problem — until it becomes a supply problem. Money and oil are bid together because a supply-shock regime pays both. That is the tape.

The trade is not 'sell everything.' The trade is: hedge the index around 740, own the AI cash-flow compounders on any real flush, do not chase SPCX or crude here, and let the next hyperscaler print or the next Hormuz headline decide whether we open 728 or reclaim 745.

Discipline over conviction. Trade the level, not the narrative.

Everything on this page is my opinion based on publicly available reporting and my read of the tape as of publication on July 23, 2026. It is not investment advice and not a recommendation to buy, sell, or short any security or derivative. Positions and levels discussed can change without notice and can invalidate the same session. Verify independently before risking capital.